255 matches found
Cisco Finds Open-Weight AI Models Easy to Exploit in Long Chats
Cisco’s new research shows that open-weight AI models, while driving innovation, face serious security risks as multi-turn attacks, including conversational persistence, can bypass safeguards and expose data...
Baohuo Android Malware Hijacks Telegram Accounts via Fake Telegram X
New Android malware Baohuo hijacks Telegram X accounts, stealing data and controlling chats. Over 58,000 devices infected, mainly in India and Brazil...
A week in security (October 6 – October 12)
Last week on Malwarebytes Labs: Apple voices concerns over age-check law that could put user privacy at risk Your passwords don’t need so many fiddly characters, NIST says Millions of very private chats exposed by two AI companion apps Fake VPN and streaming app drops malware that drains your ban...
Millions of (very) private chats exposed by two AI companion apps
Cybernews discovered how two AI companion apps, Chattee Chat and GiMe Chat, exposed millions of intimate conversations from over 400,000 users. This is not the first time we have to write about AI "girlfriends" exposing their secrets—and it probably won't be the last. This latest incident is a...
CVE-2025-60378
CVE-2025-60378 affects RISE Ultimate Project Manager & CRM. A stored HTML injection allows authenticated users to inject arbitrary HTML into invoices and messages, with injected content rendering in emails, PDFs, and messaging/chat modules sent to clients or team members. This enables phishing, c...
EUVD-2018-15773
Malware in sbrugna...
EUVD-2018-15772
Malware in sbrugna...
GHSA-7CXJ-W27X-X78Q SillyTavern Web Interface Vulnerable DNS Rebinding
Summary The web UI for SillyTavern is susceptible to DNS rebinding, allowing attackers to perform actions like install malicious extensions, read chats, inject arbitrary HTML for phishing, etc. Details DNS rebinding is a method to bypass the CORS policies by tricking the browser into resolving...
SillyTavern Web Interface Vulnerable DNS Rebinding
Summary The web UI for SillyTavern is susceptible to DNS rebinding, allowing attackers to perform actions like install malicious extensions, read chats, inject arbitrary HTML for phishing, etc. Details DNS rebinding is a method to bypass the CORS policies by tricking the browser into resolving...
EUVD-2025-6912
Malicious code in bioql PyPI...
EUVD-2025-8202
Malicious code in bioql PyPI...
CVE-2025-59422 Dify Has Broken Access Control on Log Message Endpoint Allows Reading of Chats of Others
Dify is an open-source LLM app development platform. In version 1.8.1, a broken access control vulnerability on the /console/api/apps/chat-messages?conversationid=&limit=10 endpoint allows users in the same workspace to read chat messages of other users. A regular user is able to read the query...
Leaked ChatGPT Chats: Users Treat AI as Therapist, Lawyer, Confidant
Leaked ChatGPT chats reveal users sharing sensitive data, resumes, and seeking advice on mental health, exposing risks of…...
Linux Distros Unpatched Vulnerability : CVE-2023-28686
The Linux/Unix host has one or more packages installed that are impacted by a vulnerability without a vendor supplied patch available. - Dino before 0.2.3, 0.3.x before 0.3.2, and 0.4.x before 0.4.2 allows attackers to modify the personal bookmark store via a crafted message. The attacker can...
Dark Web Roast - July 2025 Edition
Dark Web Roast - July 2025 Edition By Trellix Advanced Research Center · August 19, 2025 Executive Summary July 2025 delivered a masterclass in cybercriminal mediocrity that would make even the most charitable threat intelligence analyst weep into their coffee. After extensive hunts across the da...
Malicious code in lurk-the-chats (npm)
The package lurk-the-chats was found to contain malicious code...
Malicious code in test-mlw2-chats-loofa (npm)
The package test-mlw2-chats-loofa was found to contain malicious code...
MAL-2025-25697 Malicious code in lurk-the-chats (npm)
The package lurk-the-chats was found to contain malicious code...
MAL-2025-35062 Malicious code in test-mlw2-chats-loofa (npm)
The package test-mlw2-chats-loofa was found to contain malicious code...
CVE-2025-54868 LibreChat exposes arbitrary chats through Meilisearch engine
LibreChat is a ChatGPT clone with additional features. In versions 0.0.6 through 0.7.7-rc1, an exposed testing endpoint allows reading arbitrary chats directly from the Meilisearch engine. The endpoint /api/search/test allows for direct access to stored chats in the Meilisearch engine without...